eep

Thanks, I did have i915 drivers loaded in my go file for frigate transcoding. I've disabled that and will see how it goes.

syslogserver-diagnostics-20220401-1657.zip As title says, my server has starting becoming unresponsive and requires a hard reboot (have to hold power button down to shutdown). When unresponsive it is dropped completely from the network so can't SSH or anything. This started happening 6 to 9 months ago and has happened about 5 times. It happened today, April 1st sometime between 10am and 4:50pm (hard reboot done at 4:53). I started logging to the flash drive to try and catch any errors but I don't see any. I am wondering if it is a hardware issue as the motherboard is about 10 years old. Any insight is greatly appreciated!

Thanks for replying Jorge, while that didn't help it did motivate me to find a solution. I knew it had to be corruption form an unclean shutdown. Ran defrag and it cleared up all the "unused" space. BTW I was able to run defrag while my VM and dockers were live! btrfs filesystem defragment -rvf /mnt/cachetwo See discussion below, this seems like a big issue with BTRFS, I hope there is an alternative soon for RAID1 cache. Luckily, Unraid does not use snapshots by default so no issues with defrag adding used space.

But I am seeing 60GB of files in Samba and in Dolphin when browsing the Cache drive (including showing hidden folders). How do I see exactly whats taking up the space? I could be symlinks but not sure.

run ls -l /dev/serial/by-id/ in console For an excellent complete guide, see:

Edit: see my reply, ran defrag and space was re-allocated. I am having the same problem others have had that I am linking to below, although there didn't seem to be a known solution. I tried invoking Mover, scrub, and Check Filesystem Status in maintenance mode and it found no errors. I have 60GB of files on the cache drive according to dolphin and the du command but 97GB is being used as reported by btrfs and the GUI. I suspect this is "ghost" files from Unraid crashing while files were open but how do I find/delete those? root@Server:/mnt# du -sh /mnt/cachetwo 60G /mnt/cachetwo root@Server:/mnt# btrfs fi df -H /mnt/cachetwo Data, RAID1: total=117.85GB, used=96.96GB System, RAID1: total=33.55MB, used=16.38kB Metadata, RAID1: total=2.15GB, used=437.42MB GlobalReserve, single: total=256.61MB, used=0.00B

EDIT: 0 Day retain appears to be working now (except for a few corrupt recordings from a wifi cam), going to try 1 day retain to see if it will also work. I am on 0.9.4. The issue started with 0.9.x. Config: record: enabled: True retain_days: 0 events: max_seconds: 300 pre_capture: 5 post_capture: 5 required_zones: [] retain: default: 10 I get the following on bootup, the driveway.record ERROR repeats but there is no issues with that camera's recording files. [2022-01-11 08:39:44] frigate.record WARNING : Discarding a corrupt recording segment: Driveway-20220111083920.mp4 [2022-01-11 08:39:44] frigate.record WARNING : Discarding a corrupt recording segment: Backyard-20220111083918.mp4 [2022-01-11 08:39:44] frigate.record WARNING : Discarding a corrupt recording segment: Garage-20220111083916.mp4 [2022-01-11 08:40:59] ffmpeg.Driveway.record ERROR : [rtsp @ 0x5590c6f30b00] Could not find codec parameters for stream 0 (Video: h264, none): unspecified size [2022-01-11 08:40:59] ffmpeg.Driveway.record ERROR : Consider increasing the value for the 'analyzeduration' and 'probesize' options [2022-01-11 08:40:59] ffmpeg.Driveway.record ERROR : Output file #0 does not contain any stream [2022-01-11 08:41:39] watchdog.Driveway INFO : Terminating the existing ffmpeg process... [2022-01-11 08:41:39] watchdog.Driveway INFO : Waiting for ffmpeg to exit gracefully... Exception in thread recording_cleanup: Traceback (most recent call last): File "/usr/lib/python3.8/threading.py", line 932, in _bootstrap_inner self.run() File "/opt/frigate/frigate/record.py", line 310, in run self.expire_files() File "/opt/frigate/frigate/record.py", line 280, in expire_files oldest_timestamp = p.stat().st_mtime - 1 File "/usr/lib/python3.8/pathlib.py", line 1198, in stat return self._accessor.stat(self) FileNotFoundError: [Errno 2] No such file or directory: '/media/frigate/recordings/2022-01/01/09/Front_Yard/16.19.mp4'

I am having an issue where Frigate will not delete any recordings past the time saved in my config (I had it at 1 day and tried 0 also). I suspect this is a permission issue, as I cannot delete the recordings from explorer either, I have to delete through Unraid console or dolphin. I have tried permission fix in Unraid and completely removing and reinstalling Frigate and tried turning on Privileged Mode . However, any new recordings generating from Frigate are not deleted. I did not have this issue before the new version and clips successfully delete after the config'd 10 days.